DESCRIPTION (provided by investigator): I am committed to understanding the neurobiological underpinnings of borderline personality disorder. This disorder causes a great deal of suffering to patients and their families. Work in this area seems most productive when it focuses on a dimension of the disorder that can be measured objectively. Impulsive aggression is a particularly troubling attribute of this disorder and also forms a type of behavior that can be measured objectively. This study will use a laboratory model, the Point Subtraction Aggression Paradigm, to provoke aggression in 40 borderline patients with impulsive aggression and 40 normal controls. I will examine the differences between borderline patients and normal controls in brain activity during aggression provocation with positron emmission tomography and will focus on inhibitory regions including anterior cingulate gyrus and ventromedial orbital cortex. If borderline patients are unable to recruit inhibitory brain regions to control aggression, then this provides a very clinically meaningful explanation to patients about why they seem to struggle with self-control. It also provides a biological model that can be used to characterize a behavior that is only manifest under certain circumstances. Borderline patients can appear in many situations to be completely normal, and yet this is a disorder associated with severe functional impairment, with a high rate of suicide and with frequent acts of aggression. Because this disorder is somewhat elusive to clinicians in short-term contacts, it is especially important to provide a way of characterizing the disorder that captures the pathology clearly. In the future, I hope to use the paradigm to study treatment outcomes in borderline patients to examine the effects on regional brain control of aggression in response to both pharmacologic and psychotherapeutic interventions. Specific Aims 1) To identify and clinically characterize 40 impulsive/aggressive borderline patients (20 males, 20 females) as defined by DSM-IV and 40 normal controls (20 males, 20 females) over four years, 2) To measure regional glucose metabolic rate by Positron Emission Tomography camera (PET) scan using 18FDG as a tracer in the impulsive aggressive borderline patients and normal controls during provocation with the Point Subtraction Aggression Paradigm.